Axus Technology, a global provider of chemical mechanical planarization (CMP) equipment, critical for semiconductor and compound semiconductor fabrication, today announced its flagship Capstone CS200 platform tools offer the industry’s lowest cost of ownership (CoO) for CMP processes on 200mm silicon carbide (SiC) wafers. Compared to its closest competitor, Axus’s small-footprint Capstone delivers twice the throughput at less than half the total cost per wafer.
Yole Group forecasts the overall SiC manufacturing tool market to top US$4.4 billion by 2029. “The unique properties of SiC require specialized manufacturing tools and lines for processing power SiC devices,” the market analyst firm noted earlier this year. Axus anticipated this need, designing the state-of-the-art Capstone from the ground up to deliver advanced processing capabilities for SiC in power electronics and other applications.
“Many 200mm fabs are looking to upgrade their installed base of CMP tools to products with leading-edge capability and functionality. Our ability to deliver industry-low CoO further underscores our strong market position and capacity to support this shift,” said Axus Technology CEO Dan Trojan. “Capstone features a streamlined workflow and integrated cleaning capability, so it requires half the process steps of older CMP tools. This allows customers to greatly lower their capex investment.”
Key Capstone CoO advantages vs. competitor
- Throughput: 2.5x wafers per hour
- Power consumption: 60% lower
- DI water consumption: 80% lower
- Footprint: 45% smaller
- Capex cost per wafer: 65% lower
- Total cost per wafer: 50% lower
Another factor contributing to Capstone’s lower CoO is its built-in Process Temperature Control (PTC) technology, which enables processing at higher pressures and speeds without exceeding temperature limits of polishing pads and other sensitive components. This feature is vital for SiC and other materials with high hardness and planarization challenges that necessitate more aggressive process conditions.
Axus built its proprietary CoO model using its own system specifications, publicly available specs for competitive tools, actual consumables costs, and real-world performance data supplied by customers. The comprehensive model factors in all CoO contributors: process variables (polish time and removal rates), polishing and cleaning consumables, power and deionized (DI) water usage, system footprint, and equipment capex including cost, utilization and wafer capacity.
While the factors mentioned in the article could result in a lower COO compared to other solutions, how can anyone be sure of the algorithms that drove the stated results? This is the exact reason that SEMATECH, SEMI, and industry experts worked together in the early to mid-1990’s to develop an industry guideline on the exact approach to be used for COO calculations. That guideline is SEMI E35. There is no mention of adherence to this standard or the other related standards on reliability (E10) or subsets of COO such as OEE (E79).
If I may be allowed some self promotion, why are managers still allowing resources to be spent recreating the wheel? Commercial software solutions exist and have since SEMATECH arranged for them in 1994. Those solutions are likely a fraction of the cost of homegrown spreadsheets and have the added advantage of meeting industry standards; thus, eliminating any skepticism on the part of the fab owner with regards to how the results were calculated.